A prospective study on diagnostic accuracy of technology-enabled early detection of oral cancer and epidemiology of tobacco and other substances use in rural India

Abstract

Background

Lip and oral cavity cancer is leading cause of cancer mortality among Indian men. This study evaluated diagnostic accuracy of mobile health (mHealth) enabled screening for early detection of oral premalignant lesions or oral cancer (OPML/OC). It also described epidemiology of tobacco and other substance use and associated oral lesions in rural northern India.

Methods

A prospective study enrolled 10,101 high-risk individuals from rural settings of Varanasi district, India, between February 2021 and June 2023. Trained field workers captured habits information and oral cavity images and provided screening as suspicious or nonsuspicious on mHealth. Onsite experts and remote specialists provided clinical diagnoses. Diagnostic accuracy of mHealth-enabled screening was evaluated. A subset of 252 participants was followed to assess changes in oral lesions.

Results

Prevalence of substance use was 55.7%, with 21.4% having OPML/OC. Sensitivity of field workers and remote diagnosis for detecting OPML/OC was moderate when compared with onsite expert. Overall, interobserver agreement was substantial. During follow-up, the remote specialists identified 30 new and 13 progressive lesions with a significant decline in the red mean parameter of red, green, and blue colour ratios.

Conclusion

Although mHealth-enabled screening demonstrated lower sensitivity in detecting OPML/OC, their high specificity and expanded access to screening positions mHealth as a valuable tool for improving oral cancer screening coverage in Varanasi. This is particularly crucial given the high burden of oral cancer driven by prevalent smokeless tobacco and areca nut use and the current lack of effective population-based screening programs in this region.